Kill Process (10 page)

I partially close my laptop screen. “Daniel, you’re missing the point. The reason people would pay for PrivacyGuard is because they don’t want to be marketed to. They don’t want personalized ads. They’re asking to be left alone.”

“That’s not going to happen. O’Connelly won’t accept anything less than six bucks per user ad revenue. Talk to O’Connelly about it.”

“Ok, fine. I’ll email him right now.” I open my laptop back up and place my hand on the keyboard.

Daniel leans forward from his chair and pushes my screen back down.

I snatch my hand away, and hiss at him. My legs react involuntary, moving me toward the door, and I warily keep an eye on Daniel. I force myself back into the seat, will myself to stay there and not make a scene. As usual, Daniel doesn’t notice a thing.

“What the heck, Angie? I didn’t mean you should really talk to O’Connelly. It’s a given. We
must
increase ad revenue. Optimized ads are more than double the revenue of non-optimized ads. Isn’t there something you can do with, uh, metadata?”

Torvalds help me, there is nothing more dangerous than when a manager learns a piece of lingo. “Metadata is still data. It’s the time it was posted, who it came from, where it was posted, who liked it. Everything other than the actual content itself.”

Daniel looks hopeful.

I shake my head. “Metadata is still personalization.”

“Well, is there something we can do without personalization?”

“Sort of, maybe. There’s a grey zone.” I wave my hands at my computer. “We can disregard everything we’ve got on their profile, their likes and dislikes, topics, and friends. We could optimize on the basis of interaction with the ads themselves: which ones they click on or hover over. We could visually and geographically optimize based on IP address geolocation. It won’t be based on anything truly personal, although now we’re getting into semantics of the word ‘personal.’ Customers will see this as an invasion of their privacy. What’s going to happen when the users realize they’re paying for PrivacyGuard and they’re not getting it?”

Daniel laughs. “What’s going to happen? Nothing’s going to happen. They’ll sit there and take it.”

Adrenaline floods my system, and I stop myself short of screaming; but for once Daniel, who is oblivious to nearly everything, sees my expression and edges away from me, his fear showing in the widening of his eyes. My own breath is ragged.
Sit there and take it.
Not on my watch.

I grab my laptop and storm out of his office, slamming his door for good measure. A dozen people look up at the commotion. I ignore them all and go back to my desk. I’m too angry to sit, so I leave my laptop there and go outside.

It’s raining, but I don’t care. I walk toward the river.

Daniel, a mere cog in the organization, is a sign of things to come. He’s escalated the level of abuse of our customers. He’s right, too. A tiny handful of users will notice, and an even smaller slice will make some noise about it, while the overwhelming majority, nearly everyone, won’t do a thing. In an attempt to make sense of their universe, they’ll defend Tomo, claim Tomo’s doing what any big business will do. They’ll shift the blame to themselves, and decide if they really wanted privacy, they probably shouldn’t post anything online. Then they’ll do it anyway, because all any of us really want is to connect to other people, and Tomo is increasingly the only place they can.

I finally arrive at the metal railing along the riverfront. The water flows past slowly, dark with only a hint of green to betray the blackness. I’ve had it. I’ve just fucking had it. I lean over the railing and scream at the water with all my might, a wordless yell of rage that goes on and on.

My throat hurts afterwards. I look up and see people staring at me.

I ignore them, pull my hood up and walk down the riverfront path. And walk and walk.

*     *     *

An hour later, I’m still outside, sitting on a bench with my raincoat drawn around me.

I’ve been working my day job, chasing down abusers in my spare time, trying to maintain a relationship with Thomas, and then, after everything else is done, trying to design a new social network. It’s not possible to give everything the attention it deserves, and solving the social network problem is coming last.

I’ve freed thirty-six women from abusive relationships. I’m getting better and faster. Still, I can help maybe fifty people in a year, in the best case, and half of them will find themselves another abusive partner and end up in the same hell again.

Every time I kill someone, a little piece of my hope for the future dies. This isn’t the dream I had the first time I laid my hands on a computer. I knew then the world was going to a better place. I believed our manifest destiny was to travel to the stars, on the
Enterprise
no less, in a post-scarcity economy, a post-conflict society. I wanted to help build that place.

What have I built? Better advertising engines. Tools to help kill people. I’m literally raining death and destruction on the world.

I leave the bench and stalk over to the river again. It’s not the raging Atlantic that pounded ceaselessly against the beach where I grew up. I want that water, the water I could scream and rage at and which responded with more fury. I yell again at the water, which flows turgidly and won’t return my anger. People stare at me once more. I ignore them and walk back along the river, my feet pounding, almost running. I would run, if it could take me away from all this.

Tomo has two billion users, and it’s abusing them all. It’s not the same thing as battering a woman, not at all. Not qualitatively, not quantitatively. I’m not even sure there’s any way to compare the two. The violation of a thousand Tomo users is nothing compared to one victim of domestic abuse.

Regardless, I can’t sit there and partake in it any more. I can’t. It’s eating away at me. If there was a man out there doing what Tomo was doing, I’d kill him in an instant.

I lack the energy to fight a war on another front. I’m trying to do too much. I’m individually freeing one woman at a time and that takes all my available time and creative energy. Then I go into work, and what’s left to fight Tomo with? Nothing.

I could build an alternative to Tomo, I know I could, if I only had the time. Something has to go.

This past weekend, with the van and the problem with the onion network, was a close call. How many more times can I roll the dice and escape being caught? How many more people can I kill, each one eating away at my soul? How long until there’s nothing left of what makes me a member of the human race?

Thomas helps me be human. I can feel love and compassion when I’m with him. Certainly Thomas is not the center of my universe. I’ll never let another man define me that way. For all that, our connection is precious. If I give up killing people, I could build a real relationship with him. These occasional dates, the arm’s-length distance I keep him at, they aren’t what I want.

I shake my head. What am I even thinking about?

Give up freeing women one by one, in exchange for taking on the bigger problem of Tomo and its abuse of their own customers. Create a new online environment, one where people are truly free. Make a difference in the world like I wanted to in the first place.

No, I’m being ridiculous. Me, an ex-security penetration tester, database administrator, and data analyst. A damaged geek. How am I going to change the world?

On the other hand, Lewis, Tomo’s founder, started the company in his dorm room with nothing except a spare Linux box.

If he could do that, why can’t I destroy Tomo by building a better competitor?

I can’t go on with this quasi-life I’ve been living. It’s time for a change.

B
Y THE TIME
I return to the office, it’s late and people are filtering out. My outburst seems forgotten. I grab my computer and head home, texting Emily to see if she can meet for a drink.

“Girlfriend, I’m in Barbados for work until Thursday. Lunch on Friday?”

I never heard of anyone going to Barbados for work. I assume she’s arranged a boondoggle on the company’s dime.

Then Thomas texts me: “Dinner?”

An hour later, we’re having cheap tacos at Mi Mero Mole, a tiny place on Division. They don’t sell hard alcohol, so Thomas brought a flask of good tequila, and we sit at a sidewalk table, having sips of fiery tequila between bites of food.

We walk back to my place afterwards, nicely buzzed. I haven’t said anything about my idea. Not until I’ve spoken to Emily, whom I like to vet my plans with. At least those that don’t involve killing anyone.

“Can I stay?” he asks, arms hugging my waist, somehow safer than the claustrophobic panic I’d feel if his arms were any higher.

“Yes,
but—”
I yawn, overcome with exhaustion from the weekend, the dreadful stress of the day, and now the tequila.

He laughs. “I got it. You’re tired. I’ll grab my stuff.”

He retrieves a small bag from his trunk, and we walk into the house holding hands.

In bed, drowsy, I feel the weight of him pressing down on the mattress, hear his quiet breathing, as his warmth crosses the inches of space between us. This is nice.

*     *     *

On Tuesday I go into work and spend a couple of hours crunching data. Across the open workspace, Daniel discusses PrivacyGuard with someone. It takes only a few overheard words before I’m angry. I try to focus on my screen as the conversation keeps intruding into my thoughts. My pulse races, though, and soon I lose all ability to concentrate.

I grab my laptop and move to a small conference room where I can work without having to listen to him. I shut down chat and email clients, so there’s no way Daniel can contact me, and focus on something I started noodling on last night, the distributed-identity problem.

When a new user joins Tomo, they want to find their friends. They can search for them by name, phone number, or email. Not many search by phone number explicitly, though the mobile app works behind the scenes by accessing the contacts on your phone and comparing their metadata against that in our central database. That allows Tomo to automatically send out friend requests. We do a variation on this on computers, relying on the user’s email address book instead.

Once you’ve joined and define at least few dozen relationships, it’s easy to suggest other people you might know by looking at the relationships you have in common. Amy knows Betty and Cindy. Betty and Cindy know Daphne. It’s a good bet Amy knows Daphne, too.

Now what happens if there isn’t one social network, but many?

How can Amy find Betty and Cindy? And how can Daphne be recommended as a friend?

I’m perched on the edge of the table, staring at the whiteboard. I’ve drawn boxes and labeled them Network A, Network B, Network C, Network D. I placed Amy in A, Betty in B, and so on.

The first thing that becomes obvious is the networks need to talk to each other, using a standard protocol. Network A needs a way of asking the other networks questions and getting back answers. Because software doesn’t speak English, that means defining protocols.

In the modern day and age, there’s only one protocol that matters: JSON-encoded REST over HTTP. Ignoring all the acronyms, the basic idea is I’ve chosen a way for the services to communicate.

Now that we know
how
they talk,
what
do they speak about?

Network A has to be able to say “Do you have a user named Betty?” and the other networks need a way to reply, saying “Here’s Betty.”

It’s a little trickier, because there is likely more than one Betty. An email address or phone number may return one user, but a name could match hundreds of people.

To make matters worse, Network A doesn’t know Betty is on Network B or C or D, so it has to ask everyone.

I spend the next several hours sketching out messages for the services to exchange. On one run for coffee, I overhear Daniel ask a coworker where I am. I duck into the restroom, wait a few minutes, and then sneak back into the conference room.

It’s the end of the day by the time I’m ready to stop, and mostly everyone’s gone from the office. Software architecture drawings are scribbled all over the whiteboard walls. The centerpiece is an interaction diagram at the center showing the exact sequence of messages exchanged. This is the software equivalent of an architectural blueprint. I take photos to record it all and wipe down the boards.

The most important thing is I’ve made a discovery. It
can
be done. This crazy idea of mine is technically feasible.

Now what?

*     *     *

“I can’t make lunch today,” Emily says, leaving me a voice message on Voxer, her new favorite messaging app. “Sorry for the short notice.”

In the background, I hear the sound of arguing, and she sounds like she’s in a room full of people fighting.

I look at the empty spot across from me at the table. Guess I’ll be eating alone. My heart sinks, though not from loneliness. I need her advice.

“Short notice implies the message comes
before
the time we’re supposed to meet. This is late notice. Or, an after-the-fact apology. How about tomorrow?”

The reply comes three minutes later.

“Can’t do lunch. Dinner?”

Thomas and I are supposed to go to dinner, but I really need to see Emily. Hell, I’ll invite her to the date. If there’s drinking involved, they’ll both have a good time.

“Dinner’s good. Tasty N Sons at 6pm.”

*     *     *

Thomas and I video chatted last night, and even over the webcam video, I could see his face drop slightly when I mentioned Emily joining us.

When I promised him he could stay over, he wiggled his eyebrows playfully, and we both laughed. So it’s all good.

Thinking about getting into bed with him tonight gives me a tingle down there.

At the restaurant, I ask the hostess for a table for three, then huddle in a corner.

“Angie!”

Emily arrives with such a burst of energy loaded into that one word that people turn from all over to stare. She gives me a peck on the cheek.